CN113997955B - Track prediction method, track prediction device, electronic equipment and storage medium - Google Patents

Track prediction method, track prediction device, electronic equipment and storage medium Download PDF

Info

Publication number
CN113997955B
CN113997955B CN202111514748.XA CN202111514748A CN113997955B CN 113997955 B CN113997955 B CN 113997955B CN 202111514748 A CN202111514748 A CN 202111514748A CN 113997955 B CN113997955 B CN 113997955B
Authority
CN
China
Prior art keywords
vehicle
neighborhood
track
time sequence
current vehicle
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202111514748.XA
Other languages
Chinese (zh)
Other versions
CN113997955A (en
Inventor
刘宇杰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
FAW Group Corp
Original Assignee
FAW Group Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by FAW Group Corp filed Critical FAW Group Corp
Priority to CN202111514748.XA priority Critical patent/CN113997955B/en
Publication of CN113997955A publication Critical patent/CN113997955A/en
Application granted granted Critical
Publication of CN113997955B publication Critical patent/CN113997955B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W60/00—Drive control systems specially adapted for autonomous road vehicles
    • B60W60/001—Planning or execution of driving tasks
    • B60W60/0015—Planning or execution of driving tasks specially adapted for safety
    • B60W60/0016—Planning or execution of driving tasks specially adapted for safety of the vehicle or its occupants
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2520/00—Input parameters relating to overall vehicle dynamics
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2530/00—Input parameters relating to vehicle conditions or values, not covered by groups B60W2510/00 or B60W2520/00
    • B60W2530/201—Dimensions of vehicle
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2554/00—Input parameters relating to objects
    • B60W2554/40—Dynamic objects, e.g. animals, windblown objects
    • B60W2554/404—Characteristics

Landscapes

  • Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Human Computer Interaction (AREA)
  • Traffic Control Systems (AREA)

Abstract

The embodiment of the invention discloses a track prediction method, a track prediction device, electronic equipment and a storage medium. The method comprises the following steps: acquiring the vehicle position and the vehicle size of a current vehicle, and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle, and respectively generating time sequence input information of each vehicle; and inputting the time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle. By the technical scheme disclosed by the embodiment of the invention, the reasonability and the accuracy of track prediction in automatic driving are improved.

Description

Track prediction method, track prediction device, electronic equipment and storage medium
Technical Field
The embodiment of the invention relates to the technical field of automatic driving, in particular to a track prediction method, a track prediction device, electronic equipment and a storage medium.
Background
In a complex traffic environment, it is important how to accurately predict the future driving trajectory of a vehicle to safely and quickly drive the vehicle, not only to plan its own path, but also to dynamically plan the following driving actions according to the real-time changes of surrounding obstacles.
Because the information relied by the prior art in the prediction process is not comprehensive enough and the error exists in the used prediction algorithm, an unreasonable movement track can be predicted, and the prediction of the running track of the vehicle is further influenced, so that the safety, the reliability and the comfort of automatic driving are greatly influenced.
Disclosure of Invention
The invention provides a track prediction method, a track prediction device, electronic equipment and a storage medium, so as to improve the rationality and accuracy of track prediction in automatic driving.
In a first aspect, an embodiment of the present invention provides a track prediction method, where the method includes:
Acquiring the vehicle position and the vehicle size of a current vehicle, and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle, and respectively generating time sequence input information of each vehicle;
and inputting the time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle.
Optionally, acquiring the vehicle size of each neighboring vehicle in the neighboring range where the current vehicle is located includes:
Identifying a vehicle type of the neighborhood vehicle, and determining a vehicle size of the neighborhood vehicle based on the vehicle type; or alternatively
And acquiring the vehicle size of the neighborhood vehicle based on the laser radar.
Optionally, the vehicle dimension includes a vehicle width;
the laser radar-based acquisition of the vehicle dimensions of the neighborhood vehicle comprises:
the vehicle width of the neighborhood vehicle is acquired based on the laser radar; or alternatively
The vehicle length of the neighborhood vehicle is acquired based on the laser radar, and the vehicle width of the neighborhood vehicle is determined based on the aspect ratio of the vehicle bounding box.
Optionally, acquiring the vehicle position of each neighboring vehicle in the neighboring range where the current vehicle is located includes:
Acquiring the distance between the neighborhood vehicle and the current vehicle based on a laser radar;
and determining and obtaining the vehicle coordinates of the neighborhood vehicle based on the current coordinates of the current vehicle, the distance and the vehicle size of the neighborhood vehicle.
Optionally, the trajectory prediction model includes:
The first timing sequence processing modules are respectively used for receiving timing sequence input information of the current vehicle and each neighborhood vehicle and respectively extracting timing sequence characteristics of the received timing sequence input information;
the convolution module is respectively connected with each first time sequence processing module, and performs fusion processing on the time sequence characteristics extracted by each first time sequence processing module to obtain fusion time sequence characteristics;
The second time sequence processing module is used for receiving time sequence input information of the current vehicle and extracting time sequence characteristics of the time sequence input information of the current vehicle;
the characteristic splicing module is respectively connected with the convolution module and the second time sequence processing module and is used for carrying out characteristic splicing on the fusion time sequence characteristics and the time sequence characteristics output by the second time sequence processing module;
And the prediction module is connected with the characteristic splicing module and is used for obtaining the characteristic based on the splicing processing to obtain the track prediction information of the current vehicle.
Optionally, the prediction module includes a lateral prediction unit, a longitudinal prediction unit, and a coordinate prediction unit, and correspondingly, the track prediction information of the current vehicle includes a lateral running prediction probability distribution, a longitudinal running prediction probability distribution, and coordinate prediction information of the current vehicle.
Optionally, the training method of the track prediction model includes:
acquiring historical track data and sample input information corresponding to the historical track data;
inputting the sample input information into a track prediction model to be trained to obtain predicted track data output by the track prediction model;
obtaining a loss term based on the predicted track data and the historical track data, and obtaining a loss function based on the loss term and a constraint term, wherein the constraint term is determined based on the distance between the current vehicle and vehicles in various fields;
and carrying out iterative training on the track prediction model based on the loss function until the training condition is met, so as to obtain the track prediction model after training.
In a second aspect, an embodiment of the present invention further provides a track prediction apparatus, where the apparatus includes:
The time sequence input information generation module is used for acquiring the vehicle position and the vehicle size of the current vehicle and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle and respectively generating time sequence input information of each vehicle;
And the track prediction information generation module is used for inputting the time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle.
In a third aspect, an embodiment of the present invention further provides an electronic device, including:
one or more processors;
storage means for storing one or more programs,
The one or more programs, when executed by the one or more processors, cause the one or more processors to implement the trajectory prediction method as provided by any embodiment of the present invention.
In a fourth aspect, embodiments of the present invention further provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the trajectory prediction method provided by any of the embodiments of the present invention.
According to the technical scheme provided by the embodiment of the invention, the time sequence input information of each vehicle is respectively generated by acquiring the vehicle position and the vehicle size of the current vehicle and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle; more comprehensive vehicle related information is obtained to obtain more accurate track prediction information; further, inputting time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle; the method and the device improve the rationality and the accuracy of track prediction in automatic driving, thereby improving the safety of automatic driving.
Drawings
In order to more clearly illustrate the technical solution of the exemplary embodiments of the present invention, a brief description is given below of the drawings required for describing the embodiments. It is obvious that the drawings presented are only drawings of some of the embodiments of the invention to be described, and not all the drawings, and that other drawings can be made according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a track prediction method according to an embodiment of the present invention;
Fig. 2 is a schematic structural diagram of a vehicle enclosure according to a second embodiment of the present invention;
fig. 3 is a schematic structural diagram of a lane coordinate system according to a second embodiment of the present invention;
FIG. 4 is a schematic diagram of a track prediction model according to a second embodiment of the present invention;
fig. 5 is a flow chart of a track prediction method according to a second embodiment of the present invention;
fig. 6 is a schematic structural diagram of a trajectory prediction device according to a third embodiment of the present invention;
Fig. 7 is a schematic structural diagram of an electronic device according to a fourth embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting thereof. It should be further noted that, for convenience of description, only some, but not all of the structures related to the present invention are shown in the drawings.
Example 1
Fig. 1 is a flowchart of a track prediction method according to a first embodiment of the present invention, where the present embodiment is applicable to a situation of predicting a running track of a current vehicle; specifically, the method is more suitable for the situation that the running track of the current vehicle is predicted based on surrounding vehicle information and the current vehicle size. The method may be performed by a trajectory prediction device, which may be implemented in software and/or hardware.
Before the technical scheme provided by the embodiment of the invention is introduced, an application scene of the embodiment of the invention is introduced in an exemplary manner, and of course, the technical scheme provided by the embodiment of the invention can also be applied to other application scenes, and the embodiment does not limit the application scenes of the technical scheme. Specifically, the application scenario of the embodiment includes: in a complex traffic environment, the accuracy of the track prediction of the vehicle running and the matching degree with the actual situation directly influence the running safety and the comfort degree of passengers, so that the vehicle needs to run safely and quickly, not only to plan the path of the vehicle, but also to dynamically plan the next driving action according to the real-time change of surrounding obstacles. Therefore, how accurately to predict the future travel track of the vehicle becomes particularly important.
Because the information relied by the prior art in the prediction process is not comprehensive enough and the prediction algorithm used has errors, an unreasonable motion track can be predicted, and the prediction of the running track of the vehicle is further influenced. Specifically, some prediction methods are used for predicting according to historical tracks of vehicles, the method ignores the mutual influence among vehicles in the running process of the vehicles, when one vehicle changes the running track, the running track of surrounding vehicles is affected, and neglecting the mutual connection among the vehicles can lead to a track prediction result of the vehicles to be far different from the actual situation. Some researchers perfects based on the limitation, extracts the characteristics of the track information of different vehicles by adopting an LSTM, fuses the extracted characteristics by adopting a social pooling method, and then predicts by combining the track information of the target vehicle to obtain a future track prediction result. Both the above two methods consider the vehicle as particles when predicting the vehicle track, and do not consider the size of the vehicle itself, but in practice, the vehicle is large and is not negligible relative to the predicted track, but the current prediction method fails to consider the actual size of the vehicle in the prediction process, so that the predicted track has a large difference from the actual situation.
In order to obtain more accurate driving track prediction information, the technical scheme of the embodiment optimizes the vehicle track prediction method in the prior art, introduces a bounding box idea, expands a coordinate-based method in the traditional prediction method to a plane, considers the size of the vehicle in the prediction process, adds physical limitation to the vehicle track prediction, and can optimize the vehicle predicted track so that the predicted track can be more in line with the actual situation.
Specifically, the technical scheme provided by the embodiment of the invention respectively generates time sequence input information of each vehicle by acquiring the vehicle position and the vehicle size of the current vehicle and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle; more comprehensive vehicle related information is obtained to obtain more accurate track prediction information; further, inputting time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle; the method and the device improve the rationality and the accuracy of track prediction in automatic driving, thereby improving the safety of automatic driving.
As shown in fig. 1, the technical scheme specifically includes the following steps:
S110, acquiring the vehicle position and the vehicle size of the current vehicle, and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle, and respectively generating time sequence input information of each vehicle.
In an embodiment of the present invention, the vehicle position may be a position of the current vehicle in the lane. Optionally, a lane coordinate system is established based on each lane in which the current vehicle travels, and the vehicle position of the current vehicle can be interpreted as the position coordinate of the current vehicle in the lane coordinate system; specifically, the position coordinates include a lateral position and a longitudinal position. The vehicle size may be a specific size of the current vehicle; in the present embodiment, the vehicle size is equivalent to the vehicle width, so the acquisition of the vehicle size of the current vehicle in the present embodiment can be understood as the acquisition of the vehicle width of the current vehicle. Alternatively, the vehicle width may be obtained directly, or may be determined indirectly based on the length of the vehicle, or may be determined based on the area occupied by the current vehicle on the lane. The neighborhood range may include, but is not limited to, other lanes adjacent to the lane in which the current vehicle is located; but of course also other positions of the lane in which the current vehicle is located. Accordingly, a neighboring vehicle may be interpreted as a vehicle traveling in an adjacent other lane or in another location in the current lane. The time series input information of the vehicle may include a vehicle position and a vehicle size of the vehicle at the current time and at the previous time; where the vehicle location includes, but is not limited to, the coordinate location of the vehicle on its lane of travel.
Alternatively, the method of acquiring the vehicle size of the current vehicle may be to obtain the vehicle width of the current vehicle based on reading vehicle data stored in advance in the vehicle.
Optionally, the method for obtaining the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle comprises identifying the vehicle type of the neighborhood vehicle, and determining the vehicle size of the neighborhood vehicle based on the vehicle type; or collecting the vehicle size of the neighborhood vehicle based on the laser radar.
Wherein the vehicle types include vehicle types obtained by classifying the vehicles based on the sizes of the vehicles. For example, vehicle types include, but are not limited to, bus, sedan, off-road vehicle, van, and the like. Of course, the vehicle in the present embodiment may also be classified based on other classification criteria, and the present embodiment does not limit the classification manner and the type of the vehicle.
Specifically, identifying the vehicle size based on the vehicle type may include scanning a vehicle identification of a neighboring vehicle based on a camera of the current vehicle, determining a vehicle type of the neighboring vehicle based on the vehicle identification and a vehicle contour of the neighboring vehicle, and determining a vehicle width of the neighboring vehicle based on a pre-stored vehicle type and vehicle size vehicle database, thereby determining the vehicle size of the neighboring vehicle.
Specifically, acquiring the vehicle size of the neighborhood vehicle based on the lidar may include scanning the neighborhood vehicle based on the lidar installed by the current vehicle and determining the vehicle size of the neighborhood vehicle based on the scanning result.
Optionally, the method for determining the vehicle size of the neighboring vehicle based on the scan result may include: collecting the vehicle width of a neighborhood vehicle based on a laser radar; or collecting the vehicle length of the neighborhood vehicle based on the laser radar, and determining the vehicle width of the neighborhood vehicle based on the aspect ratio of the vehicle bounding box.
For example, if the neighboring vehicle is in front of or behind the current vehicle, the vehicle width of the neighboring vehicle may be directly scanned, and the vehicle width of the neighboring vehicle may be directly obtained based on the scanning result, thereby determining the vehicle size of the neighboring vehicle. For another example, if the neighboring vehicle and the current vehicle run in parallel, only the vehicle length of the neighboring vehicle can be obtained based on the scanning result of the laser radar, and the vehicle width of the neighboring vehicle is further determined based on the scanned vehicle length and the aspect ratio of the preset vehicle bounding box. Wherein, the vehicle bounding box may be a bounding box that replaces the vehicle size based on the length and width of the vehicle; the aspect ratio of the vehicle bounding box may be preset based on aspect ratios of various vehicles. Optionally, the aspect ratio of the vehicle bounding box can be set to be 1:2-6, namely, different aspect ratios are determined according to different vehicle types; for example, the aspect ratio of the vehicle bounding box of the bus is 1:5; for another example, the aspect ratio of the car enclosure of the car is 1:3; as shown in fig. 2, the vehicle in the figure is equivalent to three bounding boxes, the vehicle width is l, and the vehicle length can be equivalently considered to be about 3 times of the vehicle width, namely, a rectangular vehicle can be equivalent to 3 squares, and the squares are complex in calculating the distance, so that the rectangular vehicle is further equivalent to 3 circles with the same radius. Consider the center of the entire vehicle as the distance between O 2,O1 and O 2 when predicting the trajectoryEqual to the vehicle width, the radius/>, of the circle O 1,O2,O3 Of course, the aspect ratio of the vehicle bounding box is merely an alternative embodiment, and may be specifically set according to an actual vehicle, which is not limited thereto.
Specifically, obtaining the vehicle position of the current vehicle may include determining the position of the current vehicle based on a positioning device preset by the current vehicle;
specifically, acquiring the vehicle position of each neighboring vehicle in the neighboring range of the current vehicle may include acquiring a distance between the neighboring vehicle and the current vehicle based on a laser radar; further, the vehicle coordinates of the neighborhood vehicle are determined based on the current coordinates of the current vehicle, the distance and the vehicle size of the neighborhood vehicle.
Specifically, as shown in fig. 3, a lane coordinate system is pre-established based on the road where the current vehicle is located; specifically, the current position of the vehicle is taken as the origin of coordinates, the running direction of the vehicle is taken as the vertical axis, and the vertical direction of the running of the vehicle is taken as the horizontal axis. Further, based on the acquisition result of the laser radar, the distance between the neighborhood vehicle and the current vehicle is determined, and the vehicle coordinate position of the neighborhood vehicle in the same lane coordinate system is determined based on the current coordinate of the current vehicle in the lane coordinate system, the distance between the neighborhood vehicle and the current vehicle and the vehicle size of the neighborhood vehicle.
Note that, in the lane coordinate system, the current vehicle may be used as the origin of coordinates, or any other position may be used as the origin of coordinates, and the selection of the coordinate system is not limited in this embodiment. The method for acquiring the vehicle position and the vehicle size of the current vehicle and the vehicle position and the vehicle size of each neighboring vehicle in the neighborhood range of the current vehicle are just optional embodiments, and the present embodiment may also acquire the vehicle position and the vehicle size in other manners according to actual situations, which is not limited to the acquiring method.
Further, based on the method for acquiring the vehicle position and the vehicle size of the current vehicle and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle, acquiring the vehicle position and the vehicle size of each vehicle at the current moment and at each previous moment, and generating time sequence input information of each vehicle, so as to generate track prediction information of the current vehicle based on the time sequence input information of each vehicle.
Specifically, the time-series input information may be represented based on the following expression X; wherein the expression of X includes:
wherein t represents time; x represents the vehicle position and the vehicle size of the vehicle;
specifically, the expression of x includes:
where x 0 represents the lateral coordinate position of the current vehicle, y 0 represents the longitudinal coordinate position of the current vehicle, and d 0 represents the vehicle size of the current vehicle; x 1 represents the lateral coordinate position of the first neighborhood vehicle of the current vehicle, y 1 represents the longitudinal coordinate position of the first neighborhood vehicle of the current vehicle, and d 1 represents the vehicle size of the first neighborhood vehicle of the current vehicle; x n represents the lateral coordinate position of the nth neighborhood vehicle of the current vehicle, y n represents the longitudinal coordinate position of the nth neighborhood vehicle of the current vehicle, and d n represents the vehicle size of the nth neighborhood vehicle of the current vehicle.
S120, inputting time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model, and generating track prediction information of the current vehicle.
In the present embodiment, after time series input information of the current vehicle and the neighborhood vehicle is obtained based on the above embodiment, the time series input information is input into a pre-trained trajectory prediction model, and trajectory prediction information of the current vehicle is generated based on an output result of the model.
Specifically, the track prediction model in this embodiment includes a plurality of first timing processing modules, configured to receive timing input information of a current vehicle and each neighboring vehicle, and extract timing characteristics of the received timing input information respectively; the convolution module is respectively connected with each first timing sequence processing module, and performs fusion processing on the timing sequence characteristics extracted by each first timing sequence processing module to obtain fusion timing sequence characteristics; the second time sequence processing module is used for receiving time sequence input information of the current vehicle and extracting time sequence characteristics of the time sequence input information of the current vehicle; the characteristic splicing module is respectively connected with the convolution module and the second time sequence processing module and is used for carrying out characteristic splicing on the fusion time sequence characteristics and the time sequence characteristics output by the second time sequence processing module; the prediction module is connected with the characteristic splicing module and is used for obtaining the characteristic based on the splicing processing to obtain the track prediction information of the current vehicle.
Specifically, as shown in fig. 4, the first timing processing module extracts timing input information of the current vehicle and each neighboring vehicle to obtain timing characteristics of the timing input information of each vehicle, and considers images of the neighboring vehicles of the current vehicle to obtain more accurate track prediction information; wherein the first timing processing module may be comprised of at least one TCN (Temporal convolutional network, timing convolutional network) encoder. In other words, the first timing processing module in this embodiment includes a plurality of TCN encoders. In this embodiment, any one TCN encoder corresponds to any one vehicle, and is configured to receive time sequence input information using the vehicle as a current vehicle and using other vehicles as neighboring vehicles; accordingly, time sequence input information of each vehicle taking each vehicle as a current vehicle and other vehicles as neighborhood vehicles is respectively obtained based on each TCN encoder.
Further, the convolution module performs fusion processing on the time sequence characteristics of the time sequence input information of each vehicle to obtain fusion time sequence characteristics of each vehicle so as to obtain interaction relation among the vehicles.
Further, the second time sequence processing module acquires time sequence input information of the current vehicle to obtain time sequence characteristics of the time sequence input information of the current vehicle, and further deepens information characteristics of the current vehicle to obtain more accurate track prediction information;
Further, the characteristic splicing module splices the fusion time sequence characteristics of each vehicle with the time sequence characteristics of the current vehicle to obtain spliced time sequence characteristics, and predicts a track model of the current vehicle based on more characteristics to obtain more accurate track prediction information;
Further, the splicing time sequence characteristics output by the splicing module are input to the prediction module, and track prediction information of the current vehicle output by the prediction module is obtained. The prediction module comprises a transverse prediction unit, a longitudinal prediction unit and a coordinate prediction unit; accordingly, the trajectory prediction information of the current vehicle includes a lateral traveling prediction probability distribution, a longitudinal traveling prediction probability distribution, and coordinate prediction information of the current vehicle.
Specifically, the lateral prediction unit and the longitudinal prediction unit may include a lateral softmax function and a longitudinal softmax function, respectively, for outputting a lateral traveling prediction probability distribution and a longitudinal traveling prediction probability distribution of the current vehicle. The coordinate prediction unit may include a TCN decoder for outputting coordinate prediction information of the current vehicle.
Illustratively, the trajectory prediction information may be represented by Y; illustratively, the expression for Y includes: y= [ S x,Sy, x ]; wherein S x represents the lateral travel prediction probability of the current vehicle; s y represents the longitudinal running prediction probability of the current vehicle; x represents coordinate prediction information of the current vehicle; specifically, the expression of x includes:
According to the technical scheme provided by the embodiment of the invention, the time sequence input information of each vehicle is respectively generated by acquiring the vehicle position and the vehicle size of the current vehicle and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle; more comprehensive vehicle related information is obtained to obtain more accurate track prediction information; further, inputting time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle; the method and the device improve the rationality and the accuracy of track prediction in automatic driving, thereby improving the safety of automatic driving.
Example two
Fig. 5 is a flowchart of a track prediction method according to a second embodiment of the present invention, where a step of training a track prediction model in advance is added to the step of "obtaining the current vehicle position and the vehicle size" based on the above embodiments, and explanation of terms identical to or corresponding to the above embodiments is not repeated herein. Referring to fig. 5, the track prediction method provided in this embodiment includes:
S210, training a track prediction model in advance.
In the embodiment of the invention, before the track prediction model is used for predicting the track of the current vehicle, the track model needs to be trained in advance to obtain a trained track prediction model.
Optionally, the training method for the track prediction model includes: acquiring sample input information corresponding to the historical track data; inputting sample input information into a track prediction model to be trained to obtain predicted track data output by the track prediction model; obtaining a loss term based on the predicted track data and the historical track data, and obtaining a loss function based on the loss term and a constraint term, wherein the constraint term is determined based on the distance between the current vehicle and vehicles in various fields; and carrying out iterative training on the track prediction model based on the loss function until the training condition is met, and obtaining the track prediction model after training.
Specifically, the expression of the loss function of the trajectory prediction model is as follows:
Wherein P (y|x) represents a model loss function of the trajectory prediction model; θ represents a constraint term; Representing a loss term; specifically, X is historical track data, Y is predicted track coordinates, and m i is manual segmentation of an error function; θ represents a parameter of a binary gaussian distribution for each time step in the future; illustratively, the θ expression is as follows:
Wherein j represents 3 vehicle bounding boxes of the current vehicle; k represents 3 bounding boxes of a neighborhood vehicle i of the current vehicle; the values of j and k in this embodiment may be 1, 2, and 3. (x 02,y02) coordinate information ,(x01,y01)＝(x02,y02+l0),(x03,y03)＝(x02,y02-l0).l representing the current vehicle represents the vehicle width of the current vehicle; the distance between the two bounding boxes of the current vehicle may also be represented.
Specifically, if the prediction result satisfies that the distances between the three bounding box coordinates of the target vehicle and the three bounding box coordinates of the surrounding vehicle are greater than the sum of the radii of the three bounding box coordinates, the constraint term is 0, and if the prediction result is of other structures, the constraint term is 1.
Specifically, the training is repeated on the track prediction model in training based on the loss function of the embodiment until the model converges in the training sample, that is, the loss value of the model tends to zero or tends to be stable for a long time and does not change with the increase of training times, and the feature extraction model at the moment is determined to meet the training stopping condition, that is, the training of the model is completed, and the track prediction model after the training is obtained.
S220, acquiring the vehicle position and the vehicle size of the current vehicle, and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle, and respectively generating time sequence input information of each vehicle.
S230, inputting time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model, and generating track prediction information of the current vehicle.
According to the technical scheme provided by the embodiment of the invention, the time sequence input information of each vehicle is respectively generated by acquiring the vehicle position and the vehicle size of the current vehicle and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle; more comprehensive vehicle related information is obtained to obtain more accurate track prediction information; further, inputting time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle; the method and the device improve the rationality and the accuracy of track prediction in automatic driving, thereby improving the safety of automatic driving.
The following is an embodiment of a track prediction apparatus provided in the present embodiment, which belongs to the same inventive concept as the track prediction method of the above embodiments, and reference may be made to the embodiments of the track prediction method for details that are not described in detail in the embodiments of the track prediction apparatus.
Example III
Fig. 6 is a schematic structural diagram of a trajectory prediction device according to a third embodiment of the present invention, which is applicable to a situation where performance test is performed in software test. Referring to fig. 6, the specific structure of the trajectory prediction device includes: a timing input information generation module 310 and a trajectory prediction information generation module 320; wherein,
A time sequence input information generating module 310, configured to obtain a vehicle position and a vehicle size of a current vehicle, and a vehicle position and a vehicle size of each neighboring vehicle in a neighboring range where the current vehicle is located, and generate time sequence input information of each vehicle respectively;
the track prediction information generating module 320 is configured to input the time sequence input information of the current vehicle and the neighboring vehicles into a track prediction model trained in advance, and generate track prediction information of the current vehicle.
According to the technical scheme provided by the embodiment of the invention, the time sequence input information of each vehicle is respectively generated by acquiring the vehicle position and the vehicle size of the current vehicle and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle; more comprehensive vehicle related information is obtained to obtain more accurate track prediction information; further, inputting time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle; the method and the device improve the rationality and the accuracy of track prediction in automatic driving, thereby improving the safety of automatic driving.
Based on the above embodiments, the timing input information generating module 310 includes:
A vehicle size identifying unit configured to identify a vehicle type of the neighboring vehicle, and determine a vehicle size of the neighboring vehicle based on the vehicle type; or alternatively
And the vehicle size acquisition unit is used for acquiring the vehicle size of the neighborhood vehicle based on the laser radar.
On the basis of the above embodiments, the vehicle dimension includes a vehicle width;
A vehicle size recognition unit comprising:
a vehicle width acquisition subunit, configured to acquire a vehicle width of the neighboring vehicle based on the lidar; or alternatively
And the vehicle length acquisition subunit is used for acquiring the vehicle length of the neighborhood vehicle based on the laser radar and determining the vehicle width of the neighborhood vehicle based on the aspect ratio of the vehicle bounding box.
Based on the above embodiments, the timing input information generating module 310 includes:
The distance determining unit is used for acquiring the distance between the neighborhood vehicle and the current vehicle based on a laser radar;
And the vehicle coordinate determining unit is used for determining and obtaining the vehicle coordinates of the neighborhood vehicle based on the current coordinates of the current vehicle, the distance and the vehicle size of the neighborhood vehicle.
On the basis of the above embodiments, the trajectory prediction model includes:
The first timing sequence processing modules are respectively used for receiving timing sequence input information of the current vehicle and each neighborhood vehicle and respectively extracting timing sequence characteristics of the received timing sequence input information;
the convolution module is respectively connected with each first time sequence processing module, and performs fusion processing on the time sequence characteristics extracted by each first time sequence processing module to obtain fusion time sequence characteristics;
The second time sequence processing module is used for receiving time sequence input information of the current vehicle and extracting time sequence characteristics of the time sequence input information of the current vehicle;
the characteristic splicing module is respectively connected with the convolution module and the second time sequence processing module and is used for carrying out characteristic splicing on the fusion time sequence characteristics and the time sequence characteristics output by the second time sequence processing module;
And the prediction module is connected with the characteristic splicing module and is used for obtaining the characteristic based on the splicing processing to obtain the track prediction information of the current vehicle.
On the basis of the above embodiments, the prediction module includes a lateral prediction unit, a longitudinal prediction unit, and a coordinate prediction unit, and correspondingly, the track prediction information of the current vehicle includes a lateral traveling prediction probability distribution, a longitudinal traveling prediction probability distribution, and coordinate prediction information of the current vehicle.
On the basis of the above embodiments, the training method of the trajectory prediction model includes:
acquiring historical track data and sample input information corresponding to the historical track data;
inputting the sample input information into a track prediction model to be trained to obtain predicted track data output by the track prediction model;
obtaining a loss term based on the predicted track data and the historical track data, and obtaining a loss function based on the loss term and a constraint term, wherein the constraint term is determined based on the distance between the current vehicle and vehicles in various fields;
and carrying out iterative training on the track prediction model based on the loss function until the training condition is met, so as to obtain the track prediction model after training.
The track prediction device provided by the embodiment of the invention can execute the track prediction method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
It should be noted that, in the embodiment of the track prediction device, each unit and module included are only divided according to the functional logic, but not limited to the above division, so long as the corresponding functions can be implemented; in addition, the specific names of the functional units are also only for distinguishing from each other, and are not used to limit the protection scope of the present invention.
Example IV
Fig. 7 is a schematic structural diagram of an electronic device according to a fourth embodiment of the present invention. Fig. 7 illustrates a block diagram of an exemplary electronic device 12 suitable for use in implementing embodiments of the present invention. The electronic device 12 shown in fig. 7 is merely an example and should not be construed as limiting the functionality and scope of use of embodiments of the present invention.
As shown in fig. 7, the electronic device 12 is in the form of a general purpose computing electronic device. Components of the electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, a bus 18 that connects the various system components, including the system memory 28 and the processing units 16.
Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, and a local bus using any of a variety of bus architectures. By way of example, and not limitation, such architectures include Industry Standard Architecture (ISA) bus, micro channel architecture (MAC) bus, enhanced ISA bus, video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.
Electronic device 12 typically includes a variety of computer system readable media. Such media can be any available media that is accessible by electronic device 12 and includes both volatile and nonvolatile media, removable and non-removable media.
The system memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM) 30 and/or cache memory 32. The electronic device 12 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 34 may be used to read from or write to non-removable, nonvolatile magnetic media (not shown in FIG. 7, commonly referred to as a "hard disk drive"). Although not shown in fig. 7, a magnetic disk drive for reading from and writing to a removable non-volatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive for reading from or writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In such cases, each drive may be coupled to bus 18 through one or more data medium interfaces. The system memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to carry out the functions of the embodiments of the invention.
A program/utility 40 having a set (at least one) of program modules 42 may be stored in, for example, system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment. Program modules 42 generally perform the functions and/or methods of the embodiments described herein.
The electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), one or more devices that enable a user to interact with the electronic device 12, and/or any devices (e.g., network card, modem, etc.) that enable the electronic device 12 to communicate with one or more other computing devices. Such communication may occur through an input/output (I/O) interface 22. Also, the electronic device 12 may communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN) and/or a public network, such as the Internet, through a network adapter 20. As shown in fig. 7, the network adapter 20 communicates with other modules of the electronic device 12 over the bus 18. It should be appreciated that although not shown in fig. 7, other hardware and/or software modules may be used in connection with electronic device 12, including, but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, data backup storage systems, and the like.
The processing unit 16 executes various functional applications and sample data acquisition by running a program stored in the system memory 28, for example, implementing a track prediction method step provided in the present embodiment, the track prediction method includes:
Acquiring the vehicle position and the vehicle size of a current vehicle, and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle, and respectively generating time sequence input information of each vehicle;
and inputting the time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle.
Of course, those skilled in the art will appreciate that the processor may also implement the technical solution of the sample data obtaining method provided in any embodiment of the present invention.
Example five
The fifth embodiment provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements, for example, the steps of a trajectory prediction method provided by the present embodiment, the trajectory prediction method including:
Acquiring the vehicle position and the vehicle size of a current vehicle, and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle, and respectively generating time sequence input information of each vehicle;
and inputting the time sequence input information of the current vehicle and the neighborhood vehicles into a pre-trained track prediction model to generate track prediction information of the current vehicle.
The computer storage media of embodiments of the invention may take the form of any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, either in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations of the present invention may be written in one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).
It will be appreciated by those of ordinary skill in the art that the modules or steps of the invention described above may be implemented in a general purpose computing device, they may be centralized on a single computing device, or distributed over a network of computing devices, or they may alternatively be implemented in program code executable by a computer device, such that they are stored in a memory device and executed by the computing device, or they may be separately fabricated as individual integrated circuit modules, or multiple modules or steps within them may be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any specific combination of hardware and software.
Note that the above is only a preferred embodiment of the present invention and the technical principle applied. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, while the invention has been described in connection with the above embodiments, the invention is not limited to the embodiments, but may be embodied in many other equivalent forms without departing from the spirit or scope of the invention, which is set forth in the following claims.

Claims (7)

1. A track prediction method, comprising:
acquiring the vehicle position and the vehicle size of the current vehicle;
The method for acquiring the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle comprises the following steps: collecting the vehicle length of the neighborhood vehicle based on a laser radar, and determining the vehicle width of the neighborhood vehicle based on the aspect ratio of a vehicle bounding box; the neighborhood range comprises other lanes adjacent to the lane where the current vehicle is located and other positions of the lane where the current vehicle is located; the vehicle bounding box is preset based on aspect ratio of various vehicles;
Generating time sequence input information of each vehicle respectively; wherein the time sequence input information comprises the vehicle position and the vehicle size of the vehicle at the current moment and the previous moment;
inputting time sequence input information of the current vehicle and the neighborhood vehicle into a pre-trained track prediction model to generate track prediction information of the current vehicle;
the training method of the track prediction model comprises the following steps:
acquiring historical track data and sample input information corresponding to the historical track data;
inputting the sample input information into a track prediction model to be trained to obtain predicted track data output by the track prediction model;
obtaining a loss term based on the predicted track data and the historical track data, and obtaining a loss function based on the loss term and a constraint term; wherein the constraint term is determined based on the distance between the current vehicle and each neighborhood vehicle;
performing iterative training on the track prediction model based on the loss function until a training condition is met, so as to obtain a track prediction model after training is completed;
Wherein, the expression of the loss function of the track prediction model is as follows:
Wherein P (y|x) represents a loss function of the trajectory prediction model; θ represents a constraint term; Representing a loss term; x is historical track data, Y is predicted track coordinates, and m i is to segment an error function manually; Θ represents parameters of a binary gaussian distribution for each time step in the future;
Wherein, constraint term θ expression is as follows:
Wherein j represents 3 vehicle bounding boxes of the current vehicle; k represents 3 bounding boxes of a neighborhood vehicle i of the current vehicle; j and k are 1,2 and 3; (x 02,y02) represents coordinate information of the current vehicle ,(x01,y01)＝(x02,y02+l0),(x03,y03)＝(x02,y02-l0);
If the prediction result meets the requirement that the distances between the three bounding box coordinates of the target vehicle and the three bounding box coordinates of the surrounding vehicle are larger than the sum of the radii of the three bounding box coordinates, the constraint term is 0; otherwise the constraint is 1.
2. The method of claim 1, wherein the obtaining the vehicle location of each neighboring vehicle within the neighborhood of the current vehicle comprises:
acquiring the distance between the neighborhood vehicle and the current vehicle based on the laser radar;
and determining and obtaining the vehicle coordinates of the neighborhood vehicle based on the current coordinates of the current vehicle, the distance and the vehicle size of the neighborhood vehicle.
3. The method of claim 1, wherein the trajectory prediction model comprises:
The first timing sequence processing modules are respectively used for receiving timing sequence input information of the current vehicle and each neighborhood vehicle and respectively extracting timing sequence characteristics of the received timing sequence input information;
the convolution module is respectively connected with each first time sequence processing module, and performs fusion processing on the time sequence characteristics extracted by each first time sequence processing module to obtain fusion time sequence characteristics;
the second time sequence processing module is used for receiving time sequence input information of the current vehicle and extracting time sequence characteristics of the time sequence input information of the current vehicle;
the characteristic splicing module is respectively connected with the convolution module and the second time sequence processing module and is used for carrying out characteristic splicing on the fusion time sequence characteristics and the time sequence characteristics output by the second time sequence processing module;
And the prediction module is connected with the characteristic splicing module and is used for obtaining the characteristic based on the splicing processing to obtain the track prediction information of the current vehicle.
4. A method according to claim 3, wherein the prediction module comprises a lateral prediction unit, a longitudinal prediction unit and a coordinate prediction unit, and the trajectory prediction information of the current vehicle comprises a lateral travel prediction probability distribution, a longitudinal travel prediction probability distribution, and coordinate prediction information of the current vehicle, respectively.
5. A trajectory prediction device, comprising:
The time sequence input information generation module is used for acquiring the vehicle position and the vehicle size of the current vehicle and the vehicle position and the vehicle size of each neighborhood vehicle in the neighborhood range of the current vehicle and respectively generating time sequence input information of each vehicle; the neighborhood range comprises other lanes adjacent to the lane where the current vehicle is located and other positions of the lane where the current vehicle is located; the time sequence input information comprises the vehicle position and the vehicle size of the vehicle at the current moment and the previous moment;
The time sequence input information generation module comprises a vehicle size acquisition unit; the vehicle size acquisition unit comprises a vehicle length acquisition subunit;
the vehicle length acquisition subunit is used for acquiring the vehicle length of the neighborhood vehicle based on the laser radar and determining the vehicle width of the neighborhood vehicle based on the aspect ratio of the vehicle bounding box; wherein the vehicle bounding box is preset based on aspect ratio of various vehicles;
The track prediction information generation module is used for inputting time sequence input information of the current vehicle and the neighborhood vehicle into a pre-trained track prediction model to generate track prediction information of the current vehicle;
the training method of the track prediction model comprises the following steps:
acquiring historical track data and sample input information corresponding to the historical track data;
inputting the sample input information into a track prediction model to be trained to obtain predicted track data output by the track prediction model;
obtaining a loss term based on the predicted track data and the historical track data, and obtaining a loss function based on the loss term and a constraint term; wherein the constraint term is determined based on the distance between the current vehicle and each neighborhood vehicle;
performing iterative training on the track prediction model based on the loss function until a training condition is met, so as to obtain a track prediction model after training is completed;
Wherein, the expression of the loss function of the track prediction model is as follows:
Wherein P (y|x) represents a loss function of the trajectory prediction model; θ represents a constraint term; Representing a loss term; x is historical track data, Y is predicted track coordinates, and m i is to segment an error function manually; Θ represents parameters of a binary gaussian distribution for each time step in the future;
Wherein, constraint term θ expression is as follows:
Wherein j represents 3 vehicle bounding boxes of the current vehicle; k represents 3 bounding boxes of a neighborhood vehicle i of the current vehicle; j and k are 1,2 and 3; (x 02,y02) represents coordinate information of the current vehicle ,(x01,y01)＝(x02,y02+l0),(x03,y03)＝(x02,y02-l0);
If the prediction result meets the requirement that the distances between the three bounding box coordinates of the target vehicle and the three bounding box coordinates of the surrounding vehicle are larger than the sum of the radii of the three bounding box coordinates, the constraint term is 0; otherwise the constraint is 1.
6. An electronic device, comprising:
one or more processors;
storage means for storing one or more programs,
The one or more programs, when executed by the one or more processors, cause the one or more processors to implement the trajectory prediction method of any one of claims 1-4.
7. A computer readable storage medium having stored thereon a computer program, which when executed by a processor implements a trajectory prediction method as claimed in any one of claims 1 to 4.
CN202111514748.XA 2021-12-13 2021-12-13 Track prediction method, track prediction device, electronic equipment and storage medium Active CN113997955B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111514748.XA CN113997955B (en) 2021-12-13 2021-12-13 Track prediction method, track prediction device, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111514748.XA CN113997955B (en) 2021-12-13 2021-12-13 Track prediction method, track prediction device, electronic equipment and storage medium

Publications (2)

Publication Number Publication Date
CN113997955A CN113997955A (en) 2022-02-01
CN113997955B true CN113997955B (en) 2024-06-14

Family

ID=79932000

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111514748.XA Active CN113997955B (en) 2021-12-13 2021-12-13 Track prediction method, track prediction device, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN113997955B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110843789A (en) * 2019-11-19 2020-02-28 苏州智加科技有限公司 Vehicle lane change intention prediction method based on time sequence convolution network
CN111931905A (en) * 2020-07-13 2020-11-13 江苏大学 Graph convolution neural network model and vehicle track prediction method using same

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020079074A2 (en) * 2018-10-16 2020-04-23 Five AI Limited Autonomous vehicle planning
US11188082B2 (en) * 2019-01-11 2021-11-30 Zoox, Inc. Occlusion prediction and trajectory evaluation
CN110834644B (en) * 2019-10-30 2021-01-19 中国第一汽车股份有限公司 Vehicle control method and device, vehicle to be controlled and storage medium
CN113261035B (en) * 2019-12-30 2022-09-16 华为技术有限公司 Trajectory prediction method and related equipment
CN111583645B (en) * 2020-05-09 2021-08-03 北京京东乾石科技有限公司 Quality evaluation method, device, server and medium for vehicle cooperative driving

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110843789A (en) * 2019-11-19 2020-02-28 苏州智加科技有限公司 Vehicle lane change intention prediction method based on time sequence convolution network
CN111931905A (en) * 2020-07-13 2020-11-13 江苏大学 Graph convolution neural network model and vehicle track prediction method using same

Also Published As

Publication number Publication date
CN113997955A (en) 2022-02-01

Similar Documents

Publication Publication Date Title
CN110286387B (en) Obstacle detection method and device applied to automatic driving system and storage medium
CN109284348B (en) Electronic map updating method, device, equipment and storage medium
US20210302585A1 (en) Smart navigation method and system based on topological map
CN109558854B (en) Obstacle sensing method and device, electronic equipment and storage medium
CN114212110B (en) Obstacle trajectory prediction method and device, electronic equipment and storage medium
CN113554643B (en) Target detection method and device, electronic equipment and storage medium
CN113392793A (en) Method, device, equipment, storage medium and unmanned vehicle for identifying lane line
CN112712036A (en) Traffic sign recognition method and device, electronic equipment and computer storage medium
CN114419601A (en) Obstacle information determination method, obstacle information determination device, electronic device, and storage medium
CN113895460A (en) Pedestrian trajectory prediction method, device and storage medium
CN114475656B (en) Travel track prediction method, apparatus, electronic device and storage medium
Masmoudi et al. Trajectory analysis for parking lot vacancy detection system
CN114550116A (en) Object identification method and device
CN114092909A (en) Lane line extraction method and device, vehicle and storage medium
CN113591569A (en) Obstacle detection method, obstacle detection device, electronic apparatus, and storage medium
CN113997955B (en) Track prediction method, track prediction device, electronic equipment and storage medium
CN111951328A (en) Object position detection method, device, equipment and storage medium
CN113945222B (en) Road information identification method and device, electronic equipment, vehicle and medium
CN112379692B (en) Method, device and equipment for determining unmanned aerial vehicle air route and storage medium
CN115545089A (en) Point cloud data processing method, model training method, device, medium and product
CN115061386A (en) Intelligent driving automatic simulation test system and related equipment
CN114495049A (en) Method and device for identifying lane line
CN113624245B (en) Navigation method and device, computer storage medium and electronic equipment
CN113537258B (en) Action track prediction method and device, computer readable medium and electronic equipment
CN114202689A (en) Point location marking method and device, electronic equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant